1. Introduction
Soil moisture amount and its variation affect land surface sensible and latent heating fluxes. The ratio of heating and moistening of the near-surface atmosphere affects the buoyancy of air parcels that become updrafts within a boundary layer [see Pielke (2001) for a comprehensive review]. In addition, the latent heat (moisture) flux can control the lifting condensation level corresponding to the depth of the subcloud layer. The depth of the mixed boundary layer can influence the structure of thunderstorms and may also control the amount of evaporation of the precipitating particles under the cloud base.
With a cloud-resolving model, Yamada (2008) investigated the role of land surface conditions, such as soil moisture and vegetation activity, in the evolution and structure of air mass thunderstorms. The simulations with dry or wet soil moisture show clear differences in convective structure, the conditions of the subcloud layer, and the evaporation rate of precipitation within the layer. Numerical simulations also show that soil moisture can impact simulated summer rainfall in the central United States (Pan et al. 1996). An increase in soil moisture enhanced local rainfall when the lower atmosphere was thermally unstable and relatively dry, but it decreased the rainfall when the atmosphere was humid and lacked sufficient thermal forcing to initiate deep convection. Even for a small island, Hawaii Island, numerical simulations show that with improved land surface properties (soil moisture and temperature, vegetation cover), the simulated local circulations and weather are also improved (Yang et al. 2005; Yang and Chen 2008).
Mesoscale solenoidal circulations (e.g., Segal and Arritt 1992) can be generated by horizontal pressure gradients. Under hydrostatic conditions, the strength of the mesoscale circulation is proportional to the magnitude of the boundary layer virtual potential temperature horizontal gradient and the depth through which it extends (Pielke and Segal 1986). Horizontal gradients of boundary layer virtual potential temperature and boundary layer depth are attributed to differential heating over heterogeneous surfaces. Taylor et al. (2007) showed from observations that spatial variability in soil moisture and heat flux can affect the low-level wind fields on scales of 10 km upward over the Sahel. Trier et al. (2004) used a coupled high-resolution atmosphere–land surface model to show that thermodynamic stability and simulated convection initiation are affected by details in the initial soil moisture distribution through differences in the partitioning of the surface heat and moisture fluxes. The differences in convection initiation among simulations occur despite only minor differences in the overall structure of the afternoon surface moisture gradient zone.
Lying in the midlatitudes of the southwest Pacific, New Zealand consists of two main islands: the North Island and the South Island (Fig. 1). The North Island has a “spine” of mountain ranges running through the middle, with gentle rolling farmland on both sides. The central North Island is dominated by the Volcanic Plateau (with the highest mountaintop of ∼3000 m), an active volcanic area. The massive Southern Alps form the backbone of the South Island with the highest mountaintop of 3755 m. The land surface cover is mainly dominated by forests, grasses, and crops. Past studies for New Zealand suggested that the surface–atmosphere interaction is the major cause of the regional difference of weather and climate, and mesoscale phenomena (Garr and Fitzharris 1991; Sturman and Tapper 1996; etc.). In the interaction, land thermal forcing is one important factor. Revell (1984) indicated that surface convergence lines of sea breezes are regions of thunderstorm genesis (McKendry 1989, 1992; McKendry and Revell 1992; etc).
In the New Zealand limited area model (NZLAM-12), a regional atmospheric model configuration of the Met Office’s Unified Model (UM) for New Zealand with approximately a 12-km grid spacing, the soil moisture is only crudely represented in the land UM’s surface model [the Met Office Surface Exchange Scheme (MOSES-II); Cox et al. 1999; Essery et al. 2003]. For the NZLAM, the initial soil moisture is provided via an ancillary file in which the soil moisture is the monthly climate mean simulated by a GCM, lacking daily and interannual variations and therefore deviating considerably from the actual soil moisture. In New Zealand there are currently more than 50 stations with soil moisture observations (40-cm-depth soil). These observations show a typical dry soil with 10% volume soil moisture content and moist soil with 40% volume moisture content. It is not unusual that the soil moisture content changes more than 50% within a day and over distances of about 20 km. Chen and Dudhia (2001) used numerical simulations over the central American plain to show that 10% variations in soil moisture content at specific sites can result in a 100% variation in surface fluxes on a clear sunny day.
Most of the studies on the effects of land surface processes described above are based on large landmasses or continents. For the relatively small landmass of New Zealand surrounded by the southwest Pacific Ocean, an issue is how and by how much the land surface processes associated with the soil moisture variation can impact local weather and winds. Investigation of this issue was our original overall objective. However, in our analyses of simulations for the South Island with uniform soil moisture content and uniform soil type, we found that the effects of soil moisture on the surface air temperature were not uniform, implying the possible modification of this impact by mountains. Therefore, we extended our analyses to examine how and how much the effects of soil moisture content can be modified by mountains. To our knowledge, this issue has not been previously investigated in the literature.
The rest of the paper is organized as follows. A short description of the model used and the experimental setup is given in section 2; synoptic situations and the mean state of surface winds and pressure for the 6-day period are described in section 3. The effect of soil moisture content on the surface thermal fields and winds is analyzed in section 4, and modification of the effect of soil moisture by terrain is analyzed in section 5. A summary is given in the last section.
2. Model description and experimental setup
The UM has a nonhydrostatic, fully compressible, deep-atmosphere formulation using a terrain-following, height-based vertical coordinate. It employs a horizontally staggered Arakawa-C grid and a vertically staggered Charney–Phillips grid; semi-Lagrangian advection for all prognostic variables, except density, with conservative and monotone treatment of tracers; predictor–corrector implementation of a two-time-level, semi-implicit time integration scheme; and three-dimensional iterative solution of a variable-coefficient elliptic equation for the pressure increment at each time step [see Davies et al. (2005) and Webster et al. (2003) for detailed descriptions]. In all simulations, we employed 38 levels in the vertical with the model lid placed at about 39 km, with the highest vertical resolution concentrated near the ground such that 10 levels span the lowest 2 km of the atmosphere. The large-scale cloud and precipitation scheme is described in Wilson and Ballard (1999) and includes a prognostic treatment of ice microphysics. Although the convection scheme is based on Gregory and Rowntree (1990), significant modifications have been made to the basic scheme as described in Webster et al. (2003).
The coupled MOSES-II surface scheme has four soil levels, at depths of 0.1, 0.35, 1.0, and 3 m beneath the surface. Nine surface types are recognized in each grid box: broadleaf trees, needle-leaf trees, temperate grass, tropical grass, shrubs, urban, inland water, bare soil, and ice. Except for land ice, a land grid box can be made up from any mixture of the first eight surface types to represent subgrid heterogeneity. Separate surface temperatures, shortwave and longwave fluxes, sensible and latent heat fluxes, ground heat fluxes, canopy moisture contents, snow masses, and snow melt rates are computed for each of the nine surface type tiles in a grid box. Air temperature, humidity, and wind speed on atmospheric model levels above the surface and soil temperature and moisture content below the surface are treated as homogeneous across a grid box [see Essery et al. (2003) and Cox et al. (1999) for detailed descriptions]. The model domain for this study is shown in Fig. 1a with a 0.11° grid length (approximately 12 km) and 324 grid points from both west to east and south to north. At this resolution, Lake Taupo, in the central area of the North Island, is well represented as a water body and the MOSES-II surface exchange scheme is not employed. However, for the small lakes and rivers in the South Island, the inland water body is represented in grid boxes with less than 60% of water body fraction and the MOSES-II surface exchange scheme is used. Webster et al. (2008) conducted numerical studies using the UM regional model for a high wind and heavy rainfall case over New Zealand. They showed that the dynamical forcing of steep terrain and its interactions with synoptic-scale environment dominated. At high resolution (1 km) for this case, the model was capable of capturing the key features of the observed winds and heavy precipitation. However, for normal weather situations in summer, the effects of thermal forcing related to soil moisture may become important. However, studies about this issue have not been done before over New Zealand.
In this study, to investigate the effects of differences in soil moisture content on the regional model simulations in summer, we conducted two parallel idealized simulations (DRY_MNTN and MOIST_MNTN) with only differences in the soil moisture contents over New Zealand. The DRY_MNTN case had 10% initial volume soil moisture content (a typical value for dry soil in New Zealand) at all New Zealand land surface grid points (Fig. 1a), and the MOIST_MNTN case had 40% initial volume soil moisture content (a typical value for soil close to saturated) at each grid point, except in the eastern part of the North Island where the saturated soil moisture content is only 38% for the soil type in that region in the model (Fig. 1b). For both simulations the initial soil moisture content of the Australian landmass on the western side of our domain is set from an ancillary file as described in the introduction. The lateral boundary conditions are provided by the global UM at ∼40-km resolution. The NZLAM simulations from 2 to 7 January 2004 performed a 3-hourly data assimilation cycle using first guess at appropriate time (FGAT)—FGAT 3- or 3.5-dimensional variational data assimilation (3DVAR and 3.5DVAR, respectively; Lorenc et al. 2000)—that optimally combines new observations1 and the model background to generate a new analysis. Once per day at 0000 UTC (1200 New Zealand Standard Time) a long (48-h) forecast was run. The initial soil moisture content for each simulation is set as in Figs. 1a and 1b but is adjusted during the forecast for evaporation and rainfall. The hourly outputs from the 12th to the 35th hours of the integration are taken to represent the full diurnal cycle for each day. For our analysis, we assume a perfect model with correct vegetation coverage and distribution, and correct soil types and distribution.
To investigate how terrain modifies the effect of soil moisture content, we repeated the simulations with the terrain height set to 0.1 m throughout New Zealand. We call these simulations DRY_FLAT and MOIST_FLAT.
3. Synoptic situation and mean state
For the first two days (2 and 3 January 2004) of the 6-day period, most of New Zealand’s weather was dominated by the western part of a subtropical high. Northeasterly winds were found over the northern part of the North Island with northwesterly flow in all other areas (not shown). A quasi-stationary cold front with northwest-to-southeast orientation lay offshore close to the south of the South Island. From 1200 UTC 4 January to 0000 UTC 7 January, the cold front swept across the country from southwest to northeast with a northeast shift of the subtropical high (not shown). Consequently, there was northwesterly flow over the northern part of the North Island after 3 January. These synoptic patterns, throughout the 6-day period, are quite common in the New Zealand summer (e.g., Smith et al. 1991; Revell et al. 2002). The westerly–northwesterly winds are disturbed by the passage of the cold front but recover afterward. As a result, for most of the 6-day period over most of New Zealand and the sea nearby to the western (windward) side, the winds are mainly northwesterly (Fig. 2a). However, on the lee side, the winds are mainly northerly and northeasterly.
The two islands of New Zealand can significantly affect the airflow approaching and past them. The dynamically induced local wind and pressure patterns over New Zealand have been studied with observations (e.g., McCauley and Sturman 1999; McKendry et al. 1986) and numerical simulations (e.g., Revell et al. 2002; Katzfey 1995). The simulated mean surface winds and pressure patterns (Fig. 2a) are consistent with the previous studies: deceleration and splitting of the incoming northwesterly with relative high pressure (Fig. 2c) on the windward side of the South Island; low pressure (Fig. 2c) and northeasterly winds on the leeside of the islands. In addition to the dynamical forcing, the terrain also modifies the effects of soil moisture: this will be analyzed in section 5.
4. Effects of soil moisture content
In this section we will analyze the differences in the mean thermal and wind fields for the 6-day period in the boundary layer as a result of differences in soil moisture content for the DRY_MNTN and MOIST_MNTN cases.
a. Daytime regime
1) Thermal fields
During the daytime, insolation heats the land and stronger evaporative cooling is seen in the MOIST_MNTN as expected. Values of the latent heat fluxes were larger by 50 to 350 W m−2 for MOIST_MNTN than DRY_MNTN (Fig. 3a). For the mean of all land grid points at 1400 New Zealand Standard Time (NZST) for the 6-day period, the difference in the latent heat flux between DRY_MNTN and MOIST_MNTN is ∼200 W m−2 (Fig. 4c). Stronger evaporation results in a moister boundary layer (Fig. 3b) and consumes a larger part of the net downward shortwave radiation (NET-SW) flux for MOIST_MNTN, leading to relatively small surface sensible heat flux (Fig. 4b). This is consistent with summer observations of the surface energy balance for well-watered pasture with evapotranspiration the dominant energy sink (Sturman and Tapper 1996). However, for DRY_MNTN with very weak evaporative cooling (<50 W m−2; Fig. 4a), the sensible heat flux for DRY_MNTN at 1400 NZST is 50–250 W m−2 larger, with 1–5 K higher simulated surface air temperature at 1.5 m over land than for MOIST_MNTN (Figs. 3c,d and 5a).
Under the same synoptic conditions and mountain dynamic forcing, the overall mean cloud distribution for the 6-day period is rather similar for DRY_MNTN and MOIST_MNTN (Figs. 6a,b). However, a close investigation of the cloud area fraction distribution reveals some differences over land, especially for low clouds. With a moister boundary layer, the overall mean low cloud coverage over land for MOIST_MNTN is more (7%–9%) than DRY_MNTN in the afternoon (Fig. 4d). These results indicate that the moister boundary layer resulting from a wetter surface facilitates more cloud formation (mostly low cloud), consistent with some previous studies over the continents (e.g., Taylor et al. 2003; Betts and Ball 1998).
More clouds for MOIST_MNTN reduce the net downward shortwave radiation (Fig. 4c). The NET-SW reduction by more clouds in the early afternoon can be as large as 100 W m−2 on some windward areas of the North Island and part of the northern areas of the South Island with gentle slopes (Figs. 6c,d). However, the net upward longwave radiation (NET-LW) flux for MOIST_MNTN case is less than the DRY_MNTN case, with differences ranging from 10 to 70 W m−2 in the early afternoon (Fig. 7).
It is interesting to note that for the mean of all land grid points for the 6 days, the NET-LW flux differences between DRY_MNTN and MOIST_MNTN during the day are almost the same magnitude but in opposite sign to NET-SW flux differences (Fig. 4c), leaving differences in latent heat flux (ΔLH) almost the same magnitude but in opposite sign to the differences in surface sensible heat flux (ΔSH); that is, ΔLH + ΔSH ≈ 0 or ΔSH/ΔLH ≈ −1 (i.e., larger latent heat flux leads to smaller sensible heat flux). Indeed, ΔSH/ΔLH ≈ −1 represents a commonly observed land surface thermal process during soil dry-down, as shown by LeMone et al. (2007), who used aircraft and surface-flux tower data in the United States in southeast Kansas. The maximum difference between DRY_MNTN and MOIST_MNTN in surface latent heat flux occurs around 1300 NZST when the NET_SW at surface reaches the maximum, corresponding to the maximum difference in surface sensible heat flux (Fig. 4c). These results suggest that the magnitude of differences in surface evaporative cooling between DRY_MNTN and MOIST_MNTN, controlled by the magnitude of shortwave radiation reaching the ground, mainly determines the magnitude of the differences in surface air temperature.
On the land surface, the air temperature differences between DRY_MNTN and MOIST_MNTN at 1400 NZST are quite nonuniform. Small values (0.5–2 K) are found on the easternmost part of the North Island and the central and the southern windward coastal slopes of the South Island (Fig. 3c). For the easternmost areas of the North Island in the model the soil has a coarse texture with 9.5% by volume soil moisture for the critical point (at which soil moisture stress starts to restrict transpiration) and 3.3% for the wilting point. For DRY_MNTN the 10% by volume initial soil moisture in these areas is still higher than the critical point. This allows almost the same evaporation in DRY_MNTN as in MOIST_MNTN with 38% by volume soil moisture for some time (Note that the model outputs were chosen between the 12- and 35-h integrations to represent the diurnal cycle for each simulation) (Figs. 8c,d). As a result, small differences in surface latent heat flux led to small differences in surface air temperature. For other areas, the critical point and wilting point of soil moisture are 24.2% and 13.6% by volume, respectively. With 10% by volume soil moisture for DRY_MNTN lower than the wilting point, no transpiration occurs. Slight evaporation comes only from bare soil with less than 10% coverage in these areas.
For the South Island, the central and southern windward coastal regions are actually steep slopes. Under a northwesterly flow, barrier jets can be dynamically induced in the southern windward areas of the South Island (Revell et al. 2002) and enhance orographic lifting. Strong orographic lifting produced 80%–100% cloud coverage in these areas with very small differences (less than 20%) between DRY_MNTN and MOIST_MNTN (Figs. 6a,b). Consequently, the NET-SW at the surface in these areas is 200–400 W m−2, less than half that in the leeside areas (Figs. 6c,d), leading to small surface air temperature differences between DRY_MNTN and MOIST_MNTN simulated in these areas. The modification of the effects of soil moisture content by mountains is further investigated in section 5.
2) Winds
Sea breezes frequently occur in the North Island and some areas of the South Island in summer because of land–sea thermal contrast. It is expected that the surface air temperature variations due to soil moisture content variations can result in variations in local wind speed and direction.
In the early afternoon, the soil moisture content difference between DRY_MNTN and MOIST_MNTN results in 1–5-K surface air temperature difference over land. As a result, 0.5–2 m s−1 stronger onshore winds or weaker offshore winds are found close to the islands for DRY_MNTN, especially on the leeside coastal regions (Fig. 3c). Further offshore, in the open sea, wind differences are weak with random directions.
Across the central North Island (line AB in Fig. 3c) with uniform soil moisture content (Fig. 9), pronounced solenoidal circulations are simulated on the leeside associated with the dynamically induced easterly winds (Fig. 2c) enhanced by sea breezes (Figs. 9a,b). Higher air temperature is found over land, and over nearby sea on the leeside for DRY_MNTN (Fig. 9c). The downstream warm air advection is commonly found in other small tropical islands such as the Hawaiian Islands (Yang et al. 2008a,b). The stronger land–sea thermal contrast for DRY_MNTN leads to stronger sea breezes–upslope flow on both the windward and leeside areas (Fig. 9c) and, as a result, stronger thermally direct circulation cells on both sides of the mountain, with their centers at 1–1.5 km above the sea level (Fig. 9c). Across the upper central North Island (line CD in Fig. 3c; see Fig. 10), significant heating for DRY_MNTN on the windward slopes enhances the land–sea thermal contrast on the windward side and the land–land thermal contrast in the interior land area (Fig. 10c). Associated with the thermal contrasts, the wind differences between DRY_MNTN and MOIST_MNTN show two thermally forced circulation cells, one over the windward side and the other over eastern North Island where the soil type keeps the surface temperature difference between DRY_MNTN and MOIST_MNTN in the range 0–2 K (Fig. 10c), as described in previous section (contrast with Fig. 9c). On the leeside slope with ∼2-K temperature difference between the cases (Fig. 10c), the wind differences also show a thermally forced circulation cell, but it is much weaker than that in Fig. 9c with stronger heating.
When the airflow passes over central South Island, it first shows a descent over the upper leeside slope (Figs. 11a,b). This may result in adiabatic warming in the lee (Smith et al. 1991; Yang and Chen 2008). Then the airflow ascends over the middle leeside slope when encountering the upslope flow. After that, the airflow shows a second descent over the coastal region. Below the descending flow is the dynamically induced flow (Fig. 2c) enhanced by sea breezes, forming the solenoidal circulations. Significant air temperature differences between DRY_MNTN and MOIST_MNTN are found over the leeside land surface, and over the nearby sea on the leeside due to downstream warm air advection (Fig. 11c). The stronger heating for DRY_MNTN caused stronger (1–2 m s−1) sea breezes about 500 m deep. Similarly to the North Island, the wind differences (Fig. 11c) also show a large thermally forced circulation cell over the low lands and the nearby sea on the leeside due to stronger land–sea thermal contrast and stronger land surface heating. As a result, the leeside solenoidal circulations are more pronounced for DRY_MNTN (Figs. 11a,b).
For the same airflow strength and stability, the upstream blocking effect of islands is mainly determined by mountain height (e.g., Smolarkiewicz and Rotunno 1990). For the South Island, relatively high pressure associated with upstream blocking is simulated on the windward side (Fig. 2c). The incoming flow is decelerated and split horizontally and forced to ascend vertically by the high pressure before reaching the coast. For MOIST_MNTN, the forced ascent starts about 50 km offshore with a stagnant point (zero wind speed point, denoted by the black bar on the bottom on Fig. 11a) about 20 km offshore, whereas for DRY_MNTN, the forced ascent starts only about 30 km offshore (Fig. 11b). Furthermore, the stagnant point upstream found in MOIST_MNTN (Fig. 11a) is hardly found for DRY_MNTN (Fig. 11b), indicating weakening of the upstream blocking effect for DRY_MNTN. In contrast, the pronounced offshore extension of the incoming airflow ascent for the South Island was not found for the North Island with shorter mountains (see Figs. 9 and 10). Idealized numerical studies by Reisner and Smolarkiewicz (1994) for airflow past an isolated obstacle with heated surfaces show that a blocked flow regime can be changed to an unblocked flow regime by surface heating. Here, for the windward side of the South Island, stronger heating of 1–1.5 K for DRY_MNTN does not change the blocked flow regime to unblocked flow but rather weakens and/or modifies it. With the weakening of the upstream blocking effect and the onshore shift of the forced ascent, a counterclockwise circulation with its center at 1.6 km above sea level was simulated on the windward side in the wind field differences (Fig. 11c). Over the windward slope, a weak thermally direct circulation cell is also simulated as a result of the relative weak land surface heating (Fig. 11c).
During the diurnal cycle (Fig. 5), maximum wind speed differences occur in the late afternoon. At some locations in the late afternoon the differences in the zonal and meridional wind components can be as large as 4 m s−1.
The above analyses show that local wind patterns over New Zealand can be directly and indirectly created–modified by soil moisture content variations that cause variations in land surface heating. The direct effect is due to the thermal contrast between land and sea or land surfaces, the same mechanisms suggested over continents (Taylor et al. 2007; Benjamin 1986; Bossert and Cotton 1994; Wolyn and McKee 1994; and others), and the indirect effect is through the weakening of the upstream blocking (deceleration, flow splitting, and forced ascent of incoming airflow) of South Island.
b. Nighttime regime
During the night, islands mainly cool by longwave radiation loss. Evaporative cooling is almost zero for DRY_MNTN whereas a small amount of evaporative cooling occurs for MOIST_MNTN (Figs. 4a,b). With the moister boundary layer in MOIST_MNTN, the NET-LW at land surface is generally smaller (0–30 W m−2) than for DRY_MNTN. For the mean of all land grid points for the 6 days, the stronger evaporative cooling in MOIST_MNTN can be offset by less NET-LW (Fig. 4c). As a result, unlike the daytime, the overall simulated surface air temperature at night is almost the same for the MOIST_MNTN and DRY_MNTN cases (Fig. 5a).
5. Modification of the effects of soil moisture by mountains
In this section we consider the DRY_FLAT and MOIST_FLAT simulations in addition to MOIST_MNTN and DRY_MNTN in order to investigate the modification of the effects of soil moisture by mountains. We will only assess the impact on the South Island because it has a relatively regular south-southwest to north-northeast mountain ridge with clear-cut windward and leeward sides, and uniform initial soil moisture content and soil type in the model. We focus on 1400 NZST for the 6-day period for the daytime regime, when the effect of soil moisture content on the surface thermal fields is close to the largest.
The cloud coverage and distribution can be greatly affected by terrain–mountains through orographic lifting on the windward slopes and descent of airflow past mountains on the leeside slopes. For both the DRY_MNTN and MOIST_MNTN cases (Fig. 12) with mountains–terrain, the overall cloud coverage on the leeside areas is less than DRY_FLAT and MOIST_FLAT without mountains–terrain. For the 6-day mean at 1400 NZST, the differences in the cloud coverage can be as large as 40% over the leeside slopes. These cloud coverage differences cause differences in NET-SW, about 50–250 W m−2 more for DRY_MNTN (MOIST_MNTN) than for DRY_FLAT (MOIST_FLAT) (Fig. 12), especially in the central leeside slopes with higher mountains. The differences in the cloud coverage and NET-SW extend offshore downstream, indicating the effect of terrain–mountains not only on land surface but also on the nearby leeside ocean.
On the leeside, from moist flat surfaces (MOIST_FLAT) to slopes (MOIST_MNTN), the incremental change in the NET-SW due to less clouds results in a pronounced change in latent heat flux (20–150 W m−2), especially on the central leeside slopes (Fig. 13b). However, from dry flat surface (DRY_FLAT) to slopes (DRY_MNTN), there is not much difference in the latent heat flux (Fig. 13a). Pronounced increase is found for the sensible heat flux (Fig. 13c). This increase is more significant from DRY_FLAT to DRY_MNTN than from MOIST_FLAT to MOIST_MNTN (Fig. 13d).
On most of the windward sides of the South Island, from flat surfaces (DRY_FLAT or MOIST_FLAT) to slopes (DRY_MNTN or MOIST_MNTN), the NET-SW is decreased by 50–200 W m−2 for both dry and moist surfaces because of more orographic clouds for slopes (Fig. 12). The reduction in NET-SW extends offshore upstream where the island blocking causes significant deceleration and reversed flow (Fig. 2c) for cloud formation (Fig. 12). The overall reduction in the NET-SW during the daytime on the windward side from flat surfaces to slopes results in reduction in surface sensible heat flux. These features are in contrast to the leeside slopes.
To examine how the terrain modifies the surface air temperature response to change in soil moisture, we calculate and plot (Fig. 13e) a quantity called ΔTmod. We obtain this by the following steps: first subtracting MOIST_MNTN from DRY_MNTN to get ΔTMNTN (see Fig. 3c) and then doing the same for the FLAT case—that is, we subtract MOIST_FLAT from DRY_FLAT to get ΔTFLAT (not shown)—and thus ΔTmod = ΔTMNTN − ΔTFLAT. If mountains do not modify the effects of soil moisture, then ΔTmod should be 0 K. Thus, Fig. 13e can be interpreted as the modification of the effects of soil moisture by terrain. On most of the leeside areas, terrain can further increase the surface air temperature differences between a dry surface and a moist surface by 0.5–1.5 K because of more significant NET-SW for a leeside slope than a flat surface. On the windward slopes, terrain can decrease the difference by 0.5 K because of less NET-SW for a windward slope.
Along the central eastern coastal region of the South Island, ΔTmod is negative (−0.25 to −1 K), in contrast to other leeside slopes (Fig. 13e). For simulations with mountains, the dynamically induced northeasterly winds (Fig. 2a) enhanced by sea breezes are simulated toward these areas from the leeside sea, bringing in relatively cool air with small difference in ΔTMNTN (1–2.5 K) in these areas (Fig. 3c). However, for simulations with mountains, northwesterly or northerly offshore winds are simulated in these areas (Fig. 2b), allowing the high sensitivity to soil moisture to extend all the way to the coast and resulting in large differences in ΔTFLAT (2–3.5 K) and negative ΔTmod in these areas.
The above analyses show that the quite nonuniform response of the simulated surface air temperature to uniform distribution of soil moisture and soil type is mainly attributed to the mountain effects through two different mechanisms for different areas over the South Island. On the slopes, mountains lead to more orographic clouds (less NET-SW) on the windward and fewer clouds (more NET-LW) on the leeside than flat land surfaces. Thus, a dry leeside slope produces higher surface air temperature than a moist leeside slope, a difference (ΔTMNTN) larger than that for a flat surface (ΔTFLAT) with the same other conditions, and the opposite for the windward steep slopes. Along the central leeside coastal region, mountains induce reversed flow there both dynamically and thermally with low sensitivity to soil moisture (smaller ΔTMNTN) while the flat terrain with northwesterly allows the high sensitivity to soil moisture (larger ΔTFLAT) to extend all the way to the leeside coast.
6. Summary
In New Zealand, observations show that typical dry (saturated) soil moisture volume content is 10% (40%). In this study, using idealized uniform distribution of soil moisture for dry and moist land surfaces with and without mountains, we have investigated the effects of soil moisture content on the land surface air temperature (1.5 m) and winds and the modification of these effects by mountains.
As some previous studies on the continents have found, a moister surface over New Zealand facilitates a moister and cooler boundary layer and more low clouds, especially in the afternoon. The soil moisture can affect local winds over New Zealand by affecting the land surface heating. In addition, our analyses yielded some new findings:
-
Over the New Zealand land surface during the 6-day time period, the reduction in the net shortwave radiation by more clouds for MOIST_MNTN can be offset by a reduction in the net upward longwave radiation. As a result, the magnitude of the differences in the simulated surface air temperature between DRY_MNTN and MOIST_MNTN is determined by the magnitude of differences in surface evaporative cooling, which in turn is mainly affected by the amount of shortwave radiation reaching the ground during daytime.
-
For the South Island, mountains can significantly modify the effects of soil moisture content on surface thermal fields during the day through variations in cloud coverage and net downward shortwave radiation at the surface. The modification of the effects of soil moisture by terrain on surface air temperature is opposite on the leeward and steep windward slopes of the South Island because of the variation in cloud coverage due to the mountains.
-
On the central leeside coastal region of the South Island, our analyses suggest a different mechanism for the modification of soil moisture effects on surface air temperature by mountains. For this process, mountains dynamically and thermally induced onshore flow on the leeside coast, bringing in air with a lower sensitivity to soil moisture.
-
In addition to the direct thermal effect of soil moisture, local wind patterns can be indirectly created and/or modified by soil moisture variations through the weakening of the upstream blocking of the South Island for dryer soils, as shown by the weakening and onshore shift of the upstream deceleration and forced ascent of incoming airflow.
For our idealized simulations with uniform soil moisture, analysis showed that differences in soil moisture between dry and moist conditions over New Zealand can significantly affect the simulated land surface air (1.5 m) temperature with differences in the early afternoon of up to 8 K. This magnitude of difference and its distribution may or may not reflect real atmospheric sensitivities. Over New Zealand, because of large spatial variations of rainfall and land surface properties, soil moisture has large spatial variations, which may cause complicated small-scale to mesoscale circulations and local convections, as shown in previous studies on the continents (e.g., Taylor et al. 2007; Taylor and Ellis 2006; Frye and Mote 2010; Trier et al. 2008). In our proposed near future research, we will run a high-resolution model (∼1.5-km grid separation) with explicit convection using more realistic initial soil moisture settings to further investigate these problems.
Acknowledgments
This research was carried out under research collaboration SC0128 with the UK Met Office and funded by the New Zealand Foundation for Research, Science and Technology under Contract C01X081. We thank the three anonymous reviewers for their invaluable comments and suggestions in improving the manuscript.
REFERENCES
Benjamin, S. G. , 1986: Some effects of surface heating and topography on the regional severe storm environment. Part II: Two-dimensional idealized experiments. Mon. Wea. Rev., 114 , 330–343.
Betts, A. K. , and J. H. Ball , 1998: FIFE surface climate and site-average dataset 1987–89. J. Atmos. Sci., 55 , 1091–1108.
Bossert, J. E. , and W. R. Cotton , 1994: Regional-scale flows in mountainous terrain. Part II: Simplified numerical experiments. Mon. Wea. Rev., 122 , 1472–1489.
Chen, F. , and J. Dudhia , 2001: Coupling an advanced land surface–hydrology model with the Penn State–NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129 , 569–585.
Cox, P. M. , R. A. Betts , C. B. Bunton , R. L. H. Essery , P. R. Rowntree , and J. Smith , 1999: The impact of new land surface physics on the GCM simulation of climate and climate sensitivity. Climate Dyn., 15 , 183–203.
Davies, T. , M. J. P. Cullen , A. J. Malcolm , M. H. Mawson , A. Staniforth , A. A. White , and N. Wood , 2005: A new dynamical core for the Met Office’s global and regional modelling of the atmosphere. Quart. J. Roy. Meteor. Soc., 131 , 1759–1782.
Essery, R. L. H. , M. J. Best , R. A. Betts , P. M. Cox , and C. M. Taylor , 2003: Explicit representation of subgrid heterogeneity in a GCM land surface scheme. J. Hydrometeor., 4 , 530–543.
Frye, J. D. , and T. L. Mote , 2010: Convection initiation along soil moisture boundaries in the southern Great Plains. Mon. Wea. Rev., 138 , 1140–1151.
Garr, C. E. , and B. B. Fitzharris , 1991: A climate classification of New Zealand based on numerical techniques. N. Z. Geogr., 47 , 60–71.
Gregory, D. , and P. R. Rowntree , 1990: A mass flux convection scheme with representation of cloud ensemble characteristics and stability-dependent closure. Mon. Wea. Rev., 118 , 1483–1506.
Katzfey, J. J. , 1995: Simulation of extreme New Zealand precipitation events. Part II: Mechanisms of precipitation development. Mon. Wea. Rev., 123 , 755–775.
LeMone, M. A. , F. Chen , J. G. Alfieri , M. Tewari , B. Geerts , Q. Miao , R. L. Grossman , and R. L. Coulter , 2007: Influence of land cover and soil moisture on the horizontal distribution of sensible and latent heat fluxes in southeast Kansas during IHOP_2002 and CASES-97. J. Hydrometeor., 8 , 68–87.
Lorenc, A. C. , and Coauthors , 2000: The Met Office global three-dimensional variational data assimilation scheme. Quart. J. Roy. Meteor. Soc., 126 , 2991–3012.
McCauley, M. P. , and A. P. Sturman , 1999: A study of orographic blocking and barrier wind development upstream of the Southern Alps, New Zealand. Meteor. Atmos. Phys., 70 , 121–131.
McKendry, I. G. , 1989: Numerical simulation of sea breezes over the Auckland region, New Zealand—Air quality implications. Bound.-Layer Meteor., 49 , 7–22.
McKendry, I. G. , 1992: Numerical simulation of sea breeze interactions over the Auckland region, New Zealand. N. Z. J. Geol. Geophys., 35 , 9–20.
McKendry, I. G. , and C. G. Revell , 1992: Mesoscale eddy development over South Auckland—A case study. Wea. Forecasting, 7 , 134–142.
McKendry, I. G. , A. P. Sturman , and I. F. Owens , 1986: A study of interacting multi-scale wind systems, Canterbury Plains, New Zealand. Meteor. Atmos. Phys., 35 , 242–252.
Pan, Z. , E. Takle , M. Segal , and R. Turner , 1996: Influences of model parameterization schemes on the response of rainfall to soil moisture in the central United States. Mon. Wea. Rev., 124 , 1786–1802.
Pielke Sr., R. A. , 2001: Influence of the spatial distribution of vegetation and soils on the prediction of cumulus convective rainfall. Rev. Geophys., 39 , 151–177.
Pielke Sr., R. A. , and M. Segal , 1986: Mesoscale circulations forced by differential terrain heating. Mesoscale Meteorology and Forecasting, P. S. Ray, Ed., Amer. Meteor. Soc., 516–548.
Reisner, J. M. , and P. K. Smolarkiewicz , 1994: Thermally forced low Froude number flow past three-dimensional obstacles. J. Atmos. Sci., 51 , 117–133.
Revell, C. G. , 1984: Annual and diurnal variation of thunderstorms in New Zealand and outlying islands. New Zealand Meteorological Service Scientific Rep. 3, 35 pp.
Revell, M. J. , J. H. Copeland , H. R. Larsen , and D. S. Wratt , 2002: Barrier jets around the Southern Alps of New Zealand and their potential to enhance alpine rainfall. Atmos. Res., 61 , 277–298.
Segal, M. , and R. W. Arritt , 1992: Nonclassical mesoscale circulations caused by surface sensible heat-flux gradients. Bull. Amer. Meteor. Soc., 73 , 1593–1604.
Smith, R. K. , R. N. Ridley , M. A. Page , J. T. Steiner , and A. P. Sturman , 1991: Southerly changes on the east coast of New Zealand. Mon. Wea. Rev., 119 , 1259–1282.
Smolarkiewicz, P. K. , and R. Rotunno , 1990: Low Froude number flow past three-dimensional obstacles. Part II: Upwind flow reversal zone. J. Atmos. Sci., 47 , 1498–1511.
Sturman, A. P. , and N. J. Tapper , 1996: The Weather and Climate of Australia and New Zealand. Oxford University Press, 476 pp.
Taylor, C. M. , and R. J. Ellis , 2006: Satellite detection of soil moisture impacts on convection at the mesoscale. Geophys. Res. Lett., 33 , L03404. doi:10.1029/2005GL025252.
Taylor, C. M. , R. J. Ellis , D. J. Parker , R. R. Burton , and C. D. Thorncroft , 2003: Linking boundary-layer variability with convection: A case-study from JET2000. Quart. J. Roy. Meteor. Soc., 129 , 2233–2253.
Taylor, C. M. , D. J. Parker , and P. P. Harris , 2007: An observational case study of mesoscale atmospheric circulations induced by soil moisture. Geophys. Res. Lett., 34 , L15801. doi:10.1029/2007GL030572.
Trier, S. B. , F. Chen , and K. W. Manning , 2004: A study of convection initiation in a mesoscale model using high-resolution land surface initial conditions. Mon. Wea. Rev., 132 , 2954–2976.
Trier, S. B. , F. Chen , K. W. Manning , M. A. LeMone , and C. A. Davis , 2008: Sensitivity of the PBL and precipitation in 12-day simulations of warm-season convection using different land surface models and soil wetness conditions. Mon. Wea. Rev., 136 , 2321–2343.
Webster, S. , A. R. Brown , D. R. Cameron , and C. P. Jones , 2003: Improvements to the representation of orography in the Met Office Unified Model. Quart. J. Roy. Meteor. Soc., 129 , 1989–2010.
Webster, S. , M. Uddstrom , H. Oliver , and S. Vosper , 2008: A high-resolution modelling case study of a severe weather event over New Zealand. Atmos. Sci. Lett., 9 , 119–128. doi:10.1002/asl.172.
Wilson, D. R. , and S. P. Ballard , 1999: A microphysically based precipitation scheme for the Meteorological Office Unified Model. Quart. J. Roy. Meteor. Soc., 125 , 1607–1636.
Wolyn, P. G. , and T. B. McKee , 1994: The mountain–plains circulation east of a 2-km-high north–south barrier. Mon. Wea. Rev., 122 , 1490–1508.
Yamada, H. , 2008: Numerical simulations of the role of land surface conditions in the evolution and structure of summertime thunderstorms over a flat highland. Mon. Wea. Rev., 136 , 173–188.
Yang, Y. , and Y. L. Chen , 2008: Effects of terrain heights and sizes on island-scale circulations and rainfall for the island of Hawaii during HaRP. Mon. Wea. Rev., 136 , 120–146.
Yang, Y. , Y. L. Chen , and F. M. Fujioka , 2005: Numerical simulations of the island induced circulations for the island of Hawaii during HaRP. Mon. Wea. Rev., 133 , 3693–3713.
Yang, Y. , S.-P. Xie , and J. Hafner , 2008a: The thermal wake of Kauai Island: Satellite observations and numerical simulations. J. Climate, 21 , 4568–4586.
Yang, Y. , J. Ma , and S. Xie , 2008b: Observations of the trade wind wakes of Kauai and Oahu. Geophys. Res. Lett., 35 , L04807. doi:10.1029/2007GL031742.
(a) Domain of the New Zealand limited area model (NZLAM) with the initial soil moisture content by volume for the DRY_MNTN case and (b) the landmass of New Zealand with the initial soil moisture content for the MOIST_MNTN case.
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
Six-day mean surface winds (a triangle, full barb, and half barb denote 5, 1, and 0.5 m s−1, respectively (these meanings also apply to all subsequent figures) and sea level pressure (hPa, contours) for the DRY_MNTN case with (a) actual terrain and (b) flat terrain (0.1 m). (c) The difference (a) − (b).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
The differences in the surface fields between DRY_MNTN and MOIST_MNTN (DRY_MNTN − MOIST_MNTN) for (a) surface latent heat flux (W m−2), (b) water vapor mixing ratio (g kg−1), (c) surface winds (barbs) and air temperature, and (d) surface sensible heat flux (W m−2). The blue thick solid lines in (c) are the locations for the cross-sectional analyses. The thin solid lines denote terrain contours with 500-m intervals (contour intervals are the same in subsequent figures).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
The mean diurnal cycle of surface sensible heat flux, net downward shortwave flux, net downward longwave flux, and latent heat flux of all land grid points over the 6-day period for (a) DRY_MNTN and (b) MOIST_MNTN. (c) The difference between DRY_MNTN and MOIST_MNTN [(a) − (b)]. Negative values in (a) and (b) signify absorption of shortwave radiation. (d) The mean difference in the low cloud area fraction over all land grid points between MOIST_MNTN and DRY_MNTN (MOIST_MNTN − DRY_MNTN).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
Box plots of the differences (DRY_MNTN − MOIST_MNTN) in the simulated (a) land surface air temperature, (b) land surface zonal wind speed, and (c) land surface meridional wind speed during the diurnal cycle for all land grid points of New Zealand for the 6 days.
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
Six-day mean total cloud area fraction at 1400 NZST for (a) DRY_MNTN and (b) MOIST_MNTN. (c),(d) As in (a),(b), but for net downward shortwave flux (W m−2) at the surface.
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
(a),(b) As in Fig. 6, but for net upward long-wave flux (W m−2) at the surface. (c) The differences between DRY_MNTN and MOIST_MNTN, (a) − (b).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
(a),(b) As in Fig. 6, but for surface sensible heat flux. (c),(d) As in (a),(b), but for latent heat flux.
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
Cross sections at 1400 NZST along the blue solid line AB in Fig. 3c (North Island) for potential temperature (K, shading) and wind streamlines for (a) MOIST_MNTN and (b) DRY_MNTN, and (c) the differences between DRY_MNTN and MOIST_MNTN, (b) − (a).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
As in Fig. 9, but along the blue solid line CD in Fig. 3c (North Island).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
As in Fig. 9, but along the blue solid line EF in Fig. 3c (South Island). The black bar below the x axis of (a) denotes the position of the stagnant point (0 wind speed) on the windward side.
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
Six-day mean differences in total cloud area fraction at 1400 NZST between (a) DRY_MNTN and DRY_FLAT and (b) MOIST_MNTN and MOIST_FLAT. (c),(d) As in (a),(b), but for net downward shortwave radiation at the surface.
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
(a)–(d) As in Fig. 12, but for surface latent heat flux and surface sensible heat flux. (e) Surface air temperature differences (ΔTmod): (ΔTMNTN = DRY_MNTN − MOIST_MNTN) − (ΔTFLAT = DRY_MNTN-FLAT − MOIST_MNTN-FLAT).
Citation: Monthly Weather Review 139, 2; 10.1175/2010MWR3324.1
The observations include those from land stations, ships, buoys, rawinsonde, aircraft, winds from geostationary satellites, the Advanced Television and Infrared Observation Satellite (TIROS) Operational Vertical Sounder (ATOVS; NOAA-15 and -16), and Special Sensor Microwave Imager (SSM/I) wind speed.
